US Treasury Secretary Bessent met with representatives from several large language model labs in San Francisco last week, according to sources familiar with the meeting. The Treasury has not released details of the conversation, which adds to a growing list of government interactions with AI developers on economic and regulatory risks.
A closed-door conversation
The meeting was not publicly announced. Participants included multiple LLM labs, but their identities have not been disclosed. A Treasury spokesperson declined to comment on the agenda or any next steps.
Why the Treasury is paying attention
As AI models become more integrated into financial services — from automated trading to lending decisions — the Treasury is working to understand the potential for market disruption. The department has previously flagged AI-related risks in financial stability reports. The meeting in San Francisco suggests the secretary is seeking direct input from the labs themselves rather than relying solely on third-party analysis.
The labs have publicly called for consistent federal rules to avoid a patchwork of state-level regulations. The Bessent meeting could be a step toward crafting a unified regulatory approach, though no formal proposals have emerged.
No date has been set for further discussions, but the Treasury is expected to continue its outreach to technology firms as the Biden administration weighs new AI guidelines.
